-
Story
-
Resolution: Done
-
Normal
-
None
-
None
The primary problem was that a situation can happen were the hardcoded apiServerTimeout of 3 seconds could be less than the PeerTimeout (which previously defaulted to 5 seconds). This might created a race condition where one node might prematurely initiate remediation—assuming a network or peer failure—when the peer was merely slow to respond due to an overloaded API server.